Course title
6M006700,7M2700001
Statistical Signal Processing

mano kazunori Click to show questionnaire result at 2018
Course content
Recently, media information processing techniques for speech, audio, image, and language data are becoming more and more complex to provide high quality communication services. Statistical signal processing, coding, and machine learning techniques are essential to obtain sufficient results.
Purpose of class
In this class, statistical signal processing techniques for modeling, prediction, estimation, and coding are lectured. Several data processing exercises with computer programming are also required. In addition, each student will conduct a presentation and discussion on topics related to statistical signal processing.
Goals and objectives
  1. You can explain the theoretical knowledge of statistical signal processing.
  2. You can solve practical statistical data processing in computer simulation exercises.
  3. You can explain actual applications of the statistical signal processing method.
Language
English
Class schedule

Class schedule HW assignments (Including preparation and review of the class.) Amount of Time Required
1. Introduction: statistical signal processing and machine learning for media
communications.
Read the syllabus. 180minutes
Review the topics and do assignments
2. Fundamentals of statistics and probability theory Read provided materials 70minutes
Review the topics and do assignments 120minutes
3. Statistical signal processing (1) digital signal processing (filters, spectral analysis) Read provided materials 70minutes
Review the topics and do assignments 120minutes
4. Statistical signal processing (2) Linear prediction techniques for speech Read provided materials 70minutes
Review the topics and do assignments 120minutes
5. Quantization and coding in speech processing (1) Read provided materials 70minutes
Review the topics and do assignments 120minutes
6. Quantization and coding in speech processing (2) Read provided materials 70minutes
Review the topics and do assignments 120minutes
7. Statistical machine learning (1): Pattern recognition Read provided materials 70minutes
Review the topics and do assignments 120minutes
8. Statistical machine learning (2): Maximum likelihood estimation Read provided materials 70minutes
Review the topics and do assignments 120minutes
9. Statistical machine learning (3): Bayesian estimation Read provided materials 70minutes
Review the topics and do assignments 120minutes
10. Statistical machine learning (4): Deep learning Read provided materials 70minutes
Review the topics and do assignments 120minutes
11. Application of statistical signal processing (1): Speech communications Read provided materials 70minutes
Review the topics and do assignments 120minutes
12. Application of statistical signal processing (2): Communication services Read provided materials 70minutes
Review the topics and do assignments 120minutes
13. Presentation and discussion (1) Preparation for the presentation. 190minutes
14. Presentation and discussion (2) Preparation for the presentation. 190minutes
Total. - - 2650minutes
Relationship between 'Goals and Objectives' and 'Course Outcomes'

Reports of the assignments Presentations Total.
1. 30% 20% 50%
2. 20% 20%
3. 30% 30%
Total. 50% 50% -
Evaluation method and criteria
Presentations 50% (each student will give a presentation) and reports 50% (assigned during classes), for a total possible score of 100 points. A total score of 60 points or higher will be required to pass.
Textbooks and reference materials
References:
Rabiner and Schafer, Theory and applications of digital speech processing, 2011.
C. Bishop, Pattern recognition and machine learning, 2006.
Dutoit and Marques, Applied signal processing - a MATLAB-based proof of concept, 2009,
Prerequisites
Basic knowledge of signal processing (Fourier transform, filter, Z-transform, ... ) and statistics and probability theory.
Computer programming skills (either C/C++, MATLAB, Python, or R ... ) for signal processing.
Office hours and How to contact professors for questions
  • One hour after each class.
  • Any reasonable times with prior appointments. E-mails are also welcome.
Relation to the environment
Non-environment-related course
Regionally-oriented
Non-regionally-oriented course
Development of social and professional independence
  • Course that cultivates an ability for utilizing knowledge
  • Course that cultivates a basic problem-solving skills
Active-learning course
About half of the classes are interactive
Course by professor with work experience
Work experience Work experience and relevance to the course content if applicatable
Applicatable Real appliations and data processing examples are explained based on the Professor's experiences in his ICT company, especially speech processing and communication systems development experiences.
Last modified : Fri Jun 28 04:02:22 JST 2019